|
|
Security log anonymization tool focusing on artificial intelligence techniques
Šťastná, Ariela ; Jurek, Michael (oponent) ; Safonov, Yehor (vedoucí práce)
SIEM systems play a fundamental role in security monitoring. They aggregate, normalise, and filter the collected event records, which presents core tasks for applying data mining techniques. In this way, SIEMs present a great source of large amounts of normalised data. These data carry potential for achieving progress in security research, data mining, and artificial intelligence, where they could improve existing methods of investigation, make the scanning of network traffic more clear, and detect more sophisticated vectors of attack. However, one of the main problems for the use of these data is the fact that the data contained in log files are in many cases sensitive and could pose a risk to security. Due to this, the processing, as well as sharing of the data, is restricted by legislation. Considering everything that has been mentioned above, a tool for anonymization of sensitive data in log files, which works along with persisting the correlations among data was developed. The main aim of the bachelor thesis is to focus on the technical and legal level of log processing and anonymization for AI. Within the research, the analysis of the most frequently occurring data in the log files and their risk assessment was performed, resulting in the creation of categories of data, based on their sensitivity. In the work, an analysis of the present SIEM systems along with the meta keys they use is performed.
|
host ::
RSS.